Controlling crystallization pathways: a universal Na⁺/K⁺ ion switch for mesostructure engineering of zeolites

Abstract

Precise mesostructure engineering of zeolites remains a formidable challenge due to the complexity of crystallization and precursor heterogeneity. Conventional methods often rely on costly organics or destructive post-synthetics, lacking simplicity and scalability. Herein, we report a facile and universal “Na⁺/K⁺ ion switch” strategy to precisely tailor the mesostructure of zeolite beta in a seed-induced system. Merely switching Na⁺ to K⁺ without altering other parameters redirects the crystallization pathway, yielding unique mulberry-like hollow nanocrystal assemblies (beta-K) instead of conventional dense single crystals (beta-Na). Through multi-curve kinetic analysis and visual tracking, we elucidate that K⁺ fosters moderately aggregated gels that evolve into semi-crystalline nanoparticles for oriented attachment, whereas Na⁺ promotes excessive gelation leading to classical dissolution–recrystallization. This ion switch effect, synergistically modulated by inorganic alkalinity, proves universally applicable, enabling predictable “dense–loose” morphology control across diverse zeolites (ZSM-5, ZSM-11, zeolite L and mordenite). The hollow, mesopore-rich beta-K mesocrystals demonstrate superior catalytic performance in macromolecular conversion (e.g., low-density polyethylene cracking), achieving a tenfold faster rate than beta-Na due to enhanced mass transfer and acid site accessibility. This work provides a green, mechanism-driven paradigm for designing functional crystalline materials with tailored architectures.

---

Introduction

Zeolite molecular sieves, as typical crystalline microporous aluminosilicate materials, play an irreplaceable role in petroleum refining, fine chemical engineering, and environmental governance, owing to their tunable acidity, excellent hydrothermal stability, and unique shape-selectivity.^1–3 The mesostructure of these materials directly dictates the accessibility of active sites and mass transfer efficiency: dense structures ensure pore integrity for shape-selective catalysis but suffer from severe diffusion limitations for bulky molecules (e.g., polyolefins); conversely, loose and hierarchical structures enhance mass transfer and active site accessibility, yet often at the cost of reduced stability or selectivity.^4–6 Thus, precise mesostructure engineering is crucial for tailoring zeolites for targeted applications. However, the synthesis of most zeolites remains confined to a narrow “crystallization window”, which is dictated by key parameters such as system alkalinity and Si/Al ratio. Deviating from this window often leads to diminished crystallinity or impurity phases, while operating within it offers limited morphology control. For instance, conventional zeolite beta synthesis yields either large dense single crystals or ultra-small nanocrystals under different feeding conditions.^7–9 Critically, the direct and continuous regulation of its “dense–loose” mesostructures under fixed crystal size or feed composition remains a formidable challenge, significantly restricting the performance adaptability of the material across scenarios.

To overcome these limitations, several strategies have been developed for zeolite mesostructure control, which can be broadly categorized into synthetic and post-synthetic approaches. The former includes designing novel organic structure-directing agents (OSDAs)^10–14 and utilizing hard/soft mesoporous templates,^15–17 while the latter involves post-synthetic acid/alkali treatments.^18,19 However, these methods often face issues of high cost, poor recoverability of OSDAs/templates, low yields, or framework damage, raising economic and environmental concerns that hinder large-scale application. In this context, the non-classical crystallization theory, particularly “crystallization by particle attachment” (CPA),^20,21 provides a new perspective. This pathway suggests that crystal growth can proceed via the assembly of metastable precursors, such as amorphous or semi-crystalline nanoparticles,^22–26 in addition to the classical ion-mediated dissolution–recrystallization mechanism.^27,28 The prospect of manipulating the assembly of these nanoscale building-blocks offers a promising route for designing diverse mesostructures.^29–34 Nevertheless, the application of CPA-based control to zeolite beta remains less explored. Existing methods, such as those using silane coupling agents,³⁵ prefabricated special seeds,³⁶ or ball milling pretreatment,³⁷ are often limited by their complexity and technical barriers. Therefore, a simple and universally applicable methodology, based on inexpensive and readily available components, is urgently needed to precisely control the crystallization pathways of zeolite beta and related materials.

On the other hand, alkali metal ions (e.g., Na⁺ and K⁺), as key, yet economical inorganic components in zeolite synthesis systems, have recently gained increasing attention for their potential in regulating mesostructures.^38–40 Our prior studies revealed that Na⁺ and K⁺ dictate growth pathways by modulating precursor surface charge and aggregation behavior.^38,39 However, these investigations primarily focused on ZSM-5, overlooking other zeolites and, most importantly, the synergistic interplay between alkali metal ions and other synthesis parameters, such as inorganic alkalinity. A series of zeolites, represented by zeolite beta, are highly sensitive to synthesis feedstocks, and the central question remains: how do fundamental factors like alkalinity, in conjunction with the Na⁺/K⁺ switch, cooperatively regulate the evolution of building-blocks and the final mesostructures? Fortunately, our previously established “multiple crystallization curves collaborative analysis” model^38,41—which accurately captures the asynchronous evolution of long-range crystallinity, short-range order, and microporosity—provides a powerful diagnostic tool to address this question.

Based on this foundation, this study introduces a facile “Na⁺/K⁺ ion switch” strategy within a seed-induced zeolite beta synthesis system (the structure/morphology characterization of the seed are shown in Fig. S1). We aim to precisely regulate the mesostructure of zeolite beta, elucidate the underlying mechanism by applying our kinetic model, and ultimately verify the universality of this ion-mediated control across multiple zeolite frameworks (e.g., ZSM-5, ZSM-11, zeolite L, and mordenite). Meanwhile, the obtained beta mesocrystals were applied to macromolecular conversion reactions (low-density polyethylene cracking), highlighting their superior catalytic performances and clarifying the mesostructure–function relationship for zeolite catalysts.

Results and discussion

A Na⁺/K⁺ ion switch directs morphological divergence in zeolite beta

Employing a seed-induced route with a low organic template dosage (n_SiO2/n_Al2O3/n_TEA2O/n_M2O/n_H2O = 100/3.333/5/20/2000, M = K or Na), we successfully synthesized beta zeolite mesocrystals whose morphology was decisively governed by the alkali metal cation. When traditional NaOH served as the alkali source, conventional twinned super-octahedrons with relatively smooth surfaces and particle sizes of 400–600 nm were obtained (denoted as beta-Na, Fig. 1A₁), consistent with literature reports for Na⁺/seed-induced systems.^9,36 In sharp contrast, merely replacing NaOH with KOH under otherwise identical conditions induced a striking morphological transformation. As a result, mulberry-like hollow nanocrystalline assemblies were obtained, exhibiting overall sizes of 400–600 nm with attached nanocrystals measuring 20–40 nm (denoted as beta-K, Fig. 1A₂).

Fig. 1 Microscopy characterization studies of beta-Na (subscript 1, left) and beta-K (subscript 2, right). (A) SEM, (B) and (C) TEM, (D) HAADF-STEM images and (E) corresponding Si, Al and O mapping images. Inset of B: related SAED image of an individual particle.

Electron microscopy analyses unambiguously corroborate the distinct growth pathways. For beta-Na, transmission electron microscopy (TEM) and high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images reveal dense interiors and coherent *BEA-type lattice fringes across the entire particle (Fig. 1B₁–D₁ and Fig. S2A). The selected area electron diffraction (SAED) patterns display sharp and regular diffraction spot arrays (Fig. 1A₁ and Fig. S2B), confirming its single-crystalline nature. For beta-K, the images clearly show a loose, mesopore-rich interior with a central cavity of ca. 200 nm (Fig. 1B₂–D₂ and Fig. S2C), comparable to the initial seed size. Critically, the lattice fringes of adjacent nanocrystals are coherently aligned, and the SAED patterns exhibit ordered diffraction spots rather than polycrystalline rings (Fig. 1B₂ and Fig. S2D), providing direct evidence for an oriented attachment growth pathway. In addition, the disordered stacking sequence of polymorph A and polymorph B characteristic of the *BEA structure could be clearly observed from the [010] direction (Fig. 1C) and the obvious streaks elongated along the c* direction in the SEAD pattern (Fig. 1B, inset) indicating the existence of stacking faults perpendicular to the ab-plane of the framework.⁴²

The profound difference in mesostructure is further quantified via N₂-sorption analysis (Fig. 2A). The beta-Na sample exhibits a Type I isotherm, characteristic of a purely microporous structure with a low external surface area (S_ext = 39 m² g⁻¹, Table S1).⁴³ In contrast, the beta-K sample displays a mixed Type I/IV isotherm with a pronounced uptake in the mid-relative pressure region, indicating the coexistence of micropores and abundant mesopores. As a result, beta-K exhibits a substantially larger external surface area (S_ext = 163 m² g⁻¹, Table S1) and a higher total pore volume (V_total = 0.684 cm³ g⁻¹). The corresponding DFT pore size distribution (Fig. 2A, inset) confirms a broad dispersion of open mesopores,⁴³ consistent with the hollow assembly morphology observed by microscopy (Fig. 1A₂–D₂). Besides, as the Fourier transform infrared spectra shown (FT-IR, Fig. 2B), the marked enhancement of external silanol peak (3730 cm⁻¹) in beta-K over beta-Na also indicates its significantly larger external specific surface area and a loose mesopore-rich structure.⁴⁴

Fig. 2 (A) N₂-sorption isotherm of beta-K/Na, and (inset) the corresponding DFT pore size distribution based on the adsorption branch. (B) Hydroxyl region vacuum FT-IR spectra of beta-K/Na. (C) PXRD patterns of as-synthesized beta-K/Na samples. (D) NH₃-TPD profiles, (E) pyridine-IR spectra, and (F) potentiometric titration curves with 2,2,6,6-tetramethylpiperidine of the H⁺-type beta-K/Na samples.

A suite of characterization techniques confirms that the “Na⁺/K⁺ ion switch” alters the mesostructure without affecting the intrinsic framework properties. Both beta-K and beta-Na samples are phase-pure *BEA structures (Fig. 2C), and possess identical micropore size (0.71 nm, Fig. S3A), similar micropore volumes/surface area (Table S1), and framework Si/Al ratios (Table S2). ²⁷Al and ²⁹Si magic-angle spinning nuclear magnetic resonance (MAS NMR) curves are virtually identical, showing Al predominantly in tetrahedral coordination (Fig. S3B) and similar Q⁴/Q³ ratios (Fig. S3C).³⁷ Besides, elemental mapping confirms a homogeneous distribution of Si and Al in both materials (Fig. 1E). Collectively, these results demonstrate that the “Na⁺/K⁺ ion switch” strategy specifically tailors the mesocrystal morphology and porosity, while leaving the framework characteristics, microporosity and chemical composition of zeolite beta unaffected.

Crystallization kinetics reveal contrasting assembly pathways

To elucidate how alkali metal cations direct the growth of zeolite beta, we interrogated a series of crystallization intermediates (beta-K/Na-t) using our multiple crystallization curves analysis model. The evolution of relative crystallinity (RC) derived from powder X-ray diffraction analysis (PXRD, Fig. S4) reveals starkly divergent kinetics between the two systems (Fig. 3A). In the Na⁺ system, crystallization commenced earlier but progressed slowly, reaching only ca. 40% RC after 28 h before accelerating. In contrast, the K⁺ system exhibited a prolonged induction period with RC below ca. 14% within the first 20 h, followed by a remarkably accelerated crystallization, achieving near-complete crystallinity within the next 12 h. This kinetic disparity strongly suggests that the underlying building-blocks and their assembly pathways differ fundamentally. Besides, the inductively coupled plasma-atomic emission spectroscopy analysis (ICP-AES, Fig. S5) further indicates that differences in soluble Si and Al species primarily occur during the early stages (t ≤ 4 h, which will be discussed later), with the compositions converging as crystallization proceeds.

Fig. 3 (A) The kinetic crystal growth curve according to PXRD relative intensity. The evolution of short-range order (via UV-Raman, red line) and measurable micropore volume (by N₂-sorption, blue line) vs. the change of RC for (B) beta-K and (C) beta-Na.

Given the distinct XRD growth profiles, we quantified the evolution of characteristic micropore volume (V_micro by N₂-sorption, Fig. S6/Table S3) and short-range order (by UV-Raman, Fig. S7) to deconvolute the crystallization process. A synergistic analysis of these parameters with RC uncovered two contrasting assembly mechanisms (Fig. 3B and C, detailed calculation methods shown in the SI). In the K⁺ system (Fig. 3B), a substantial level of short-range order (ca. 67%) was established long before significant long-range crystallinity (RC ≥ 15%) was detected. Notably, the development of V_micro progressed almost synchronously with the RC. This indicates that well-ordered, semi-crystalline building-blocks form at an early stage. These blocks subsequently attach to the growing crystals and, through oriented assembly, rapidly establish the microporous framework (t = 20–30 h, Fig. 3A). In the Na⁺ system (Fig. 3C), the growth of long-range order (RC) proceeded almost synchronously with short-range order, along with the V_micro developing slightly behind them throughout the main crystallization stage (RC = 14–85%). This suggests that the initially attached material is more disordered, and the open microporous network is only perfected after a delayed internal reorganization or through a dissolution–recrystallization process. Strikingly, the asynchronous trends observed here for zeolite beta are opposite to those we previously reported for ZSM-5,³⁸ highlighting a system-dependent response to alkali metal ions and raising critical questions about the universal governing principles.

Visualizing the ion-dependent evolution from precursors to crystals

The crystallization pathways for beta-Na/K diverge from the initial gel state. Dynamic light scattering characterization (DLS, Fig. S8A) and scanning electron microscopy (SEM) images of the aged gels (beta-K/Na-0) reveal that while beta seeds remain stable, of ca. 220 nm, the primary silica particles (ca. 15 nm) exhibit cation-dependent aggregation. In the K⁺ system, these particles form moderate-sized, loose clusters (ca. 200 nm, Fig. S8A and S9A), whereas in the Na⁺ system, they aggregate into large, dense lumps (>1.0 µm, Fig. S8A and S10A). This divergence is rationalized by zeta potential measurements (Fig. S8B): colloidal particles in the K⁺ system possess a more negative surface potential (−21.2 mV) than those in the Na⁺ system (−8.2 mV). The weaker screening effect of the larger K⁺ ion enhances electrostatic repulsion between colloidal particles. As a result, excessive gelation is suppressed, favoring the formation of looser precursor aggregates.^38,45 Consequently, the liquid phase in the K⁺ system retains an order of magnitude higher concentration of soluble silica (1217 vs. 105 mmol L⁻¹ in Na⁺ system, Fig. S5), consistent with its reduced gelation tendency.

Upon hydrothermal treatment, precursors in both the K⁺ system (Fig. 4A and Fig. S9B) and Na⁺ system (Fig. 5A and B and Fig. S10B and C) restructured into worm-like particles (WLPs) that further cross-linked. Notably, unlike our previous observations in ZSM-5 synthesis,³⁸ WLPs in both beta-K/Na-t systems aggregated into larger lumps (Fig. S9C–G and S10D–G). To quantify the degree of cross-linking, we analyzed the mesopore volume (V_meso) of low-crystallinity intermediates (RC ≤ 15%, as t ≤ 20 h for the K⁺ system and t ≤ 16 h for the Na⁺ system), which primarily reflects the disordered packing of amorphous precursors at this stage. The V_meso of beta-Na-t decreased more significantly than that of beta-K-t (Table S3), indicating a higher degree of cross-linking and fusion in the Na⁺ system, consistent with its less negative surface charge. Time-resolved DLS measurements further support these observations. In the K⁺ system (Fig. S11A), the hydrodynamic particle size increased gradually during the early stages and reached approximately 4–5 µm at intermediate times, before partially decreasing to ca. 700–1000 nm at longer times. In contrast, the Na⁺ system exhibited more rapid growth of larger aggregates (Fig. S11B), exceeding 10 µm at intermediate stages, consistent with its higher degree of cross-linking and “excessive gelation” behavior. Besides, optical photographs of centrifuged intermediates (Fig. S12) also confirm that the aggregates in the K⁺ system are more loosely packed than in the Na⁺ system.

Fig. 4 TEM images and the corresponding SAED/FFT analysis results of beta-K-t, where t = (A) 2, (B) 8, (C) 16 and (D) 24 h, and the white dashed boxes mark the ranges for SAED or FFT analysis.

Fig. 5 TEM images and the corresponding SAED patterns of beta-Na-t, where t = (A) 2, (B) 4, (C) 16, (D) 24, (E) 32 and (F) 36 h. The white dashed boxes mark the ranges for SAED analysis, the light-yellow dashed boxes indicate the positions of stepwise magnification, and the deep-yellow curves outline the smooth contour of the crystallized portion.

For the beta-K-t system, the gel fully coated the seeds and developed surface roughness by 4–8 h (Fig. 4B and Fig. S9C and D), concomitant with a rise in the short-range order to 26.3% (Fig. 3B). While the seeds retained high crystallinity, the deposited WLPs were still long-range disordered (SAED in Fig. 4B). A critical transition occurred by 16 h (RC = 7.7%): the gel coating roughened and became partially fragmented, and lattice fringes assignable to the *BEA structure were identified via FFT analysis on gel-derived particles (Fig. 4C₁–C₃ and Fig. S9E). The widespread presence of such similar semi-crystalline nanoparticles (Fig. S13) and the concurrent surge in relative short-range order to 39.8% (Fig. 3B) visually confirm the early self-organization of building-blocks inferred from kinetics. The small-angle X-ray scattering (SAXS) analysis further corroborates this evolution (Fig. S14A): at 12–20 h, the emergence of 6–8 nm semi-crystalline nanoparticles was detected, consistent with the building-blocks that was observed via TEM in Fig. S13. Besides, the synchronous but delayed increase of V_micro and RC (Fig. 3B) relative to the short-range order implies that these semi-ordered blocks must undergo CPA on the seed surface to complete the long-range micropore frameworks/networks. By 24 h, these nanoparticles assembled on seeds and enhanced their alignment, initially forming a mulberry-like morphology (Fig. S9F–G). SAED patterns showed sharp diffraction spots, evidencing the oriented attachment of these semi-ordered blocks (Fig. 4D). In addition, the concomitant presence of weak diffraction rings in residual colloidal indicates the continuous generation of *BEA short-range order in the gel (relative Raman intensity = 83.9%, Fig. 3B). By 32 h, amorphous agglomerates were largely consumed (RC = 88.8%), yielding particles resembling the final beta-K products (Fig. S9H).

The hallmark hollow structure of beta-K evolved during the subsequent ripening stage (44–60 h). Although V_micro reached 82% of its final value by 32 h, the relative V_meso was only 41.8% (Table S3), with the central seed region remaining dense (Fig. 4D). The marked increase in V_meso during ripening reflects secondary mesopore formation via partial dissolution of the Si-rich seed core (Si/Al = 17.76). Such a process is also facilitated by the high interfacial energy of CPA-generated boundaries and the lack of a protective dense Al-rich shell around the seed, allowing for selective etching in the alkaline medium, ultimately yielding the hollow architecture (which could be tracked by adsorption analysis, Fig. S6B).^18–20

In contrast, the beta-Na-t was characterized by stronger gelation, which led to denser amorphous lumps, as confirmed by the smaller stacked mesopore data (V_meso, Table S3). At the early stage (t = 4 h), amorphous WLPs self-aggregate into large agglomerates, whereas most seeds remained isolated with limited attachment (Fig. 5B and Fig. S10C). From 12–16 h, seeds became embedded in these agglomerates and showed initial faceted growth (Fig. S10D–E). Critically, the surrounding gel phase remained SAED-amorphous (Fig. 5C₁), consistent with the Raman data; while the growing crystals developed sharp interfaces with the gel (Fig. 5C₂–C₄). This suggests that crystal growth proceeded via both direct seed-induced transformation of attached gels and indirect dissolution–recrystallization of amorphous lumps. By 24–32 h, well-faceted crystals protruded from the gel aggregates (Fig. 5D and E and Fig. S10F–G), with the gel remaining SAED-amorphous (Fig. 5D₁). After heating for 32 h, numerous independent crystals appeared, with similar morphology but slightly smaller size than the final products (Fig. 5E and Fig. S10G). This predominant classical growth mode, involving the dissolution of disordered gels and deposition of soluble species, accounts for the almost synchronous development of the short- and long-range order of the *BEA structure (Fig. 3C) and the observed lag of V_micro behind RC. Besides, SAXS analysis further supports this classical crystallization pathway (Fig. S14B): no newly formed nanoparticles were detected in the Na⁺ system at 12–16 h, indicating that crystallization proceeds mainly via seed-induced growth and dissolution–recrystallization of amorphous gel lumps, without discrete particle-type building-blocks.

During the ripening stage in the Na⁺ system (t = 40–60 h), the amorphous gel agglomerates were further consumed and the beta crystals approached their final morphology (Fig. 5F and Fig. S10H), as RC reached 96.4% at 40 h (Fig. 3A). In stark contrast to the K⁺ system, the V_meso of beta-Na-t decreases (Table S3), indicating an absence of constructive etching. This is attributed to (i) fewer high-energy defect sites due to the predominance of classical growth, and (ii) the presence of a dense, protective Al-rich shell that shields the central seed from dissolution.

Mechanism of the ion switch and its synergy with system alkalinity

To directly corroborate the nature of the building-blocks in the two crystallization pathways, we performed alkaline etching on the calcined beta-K and beta-Na samples. Zeolites are preferentially etched at Si-rich regions and high-energy sites such as defects, splicing interfaces, or twin boundaries.^46,47 Strikingly, etching of beta-K fragmented its outer architecture, liberating abundant nanocrystals (Fig. S15A–C). This provides direct, visual proof that its mulberry-like morphology originates from the oriented attachment of semi-crystalline nanoparticles. In contrast, beta-Na largely retained its dense morphology after etching, with dissolution primarily confined to the Si-rich seed core and minor traces along twin interfaces (Fig. S15D–F). This supports the conclusion that its outer shell layer is predominantly constructed via the addition of simple soluble species, consistent with a classical dissolution–recrystallization pathway.

Control experiments confirmed the roles of seeds and TEAOH. In the absence of seeds, no crystallization occurred in either system (Fig. S16A), and the gels in the Na⁺ system showed stronger aggregation (Fig. S16B–E). When seeds were added without TEAOH, crystallization was impeded and phase transformation was observed (Fig. S16F–J). Thus, while seeds and TEAOH are essential for the formation and stabilization of the *BEA structure, the “Na⁺/K⁺ ion switch” is the decisive factor governing the properties of gel precursors, the evolution of building-blocks, and the mesostructure of final products.

Accordingly, we propose the crystallization mechanism outlined in Scheme 1, involving the formation of precursor gels, the evolution of building-blocks, and their attachment to the seed surface. In the K⁺ system (Scheme 1A), WLPs exhibit moderate aggregation (i–ii), which limits early attachment but allows sufficient time for microstructural ordering (iii). The resulting semi-crystalline nanoparticles then grow as building-blocks through oriented attachment on the seed surface, forming nanocrystalline assemblies (iv). And during the ripening stage, the Si-rich seeds are etched and dissolved, further leading to the formation of hollow structures (v). In the Na⁺ system (Scheme 1B), the excessive cross-linking and self-polymerization of the gel reduce its direct attachment to the seeds (i–ii). Limited contact and mass transfer hinder the structural ordering of these dense gels, whether it is seed-induced or spontaneous (iii). As a result, crystal growth proceeds mainly through dissolution–recrystallization of simple soluble species, producing dense single crystals (iv–v).

Scheme 1 Crystallization pathways of (A) beta-K and (B) beta-Na in this seed-induced system.

However, such mechanisms are not entirely consistent with our earlier findings in the ZSM-5 system,^38,39 where K⁺ promoted random attachment of amorphous particles and Na⁺ facilitated the oriented attachment of semi-crystalline nanoparticles—a reversal of the roles observed here for zeolite beta. This discrepancy underscores that the ion-specific effect is system-dependent and hints at a more complex, underlying governing principle.

To resolve this dichotomy, we identified a major difference in the intrinsic inorganic alkalinity (M₂O/H₂O feed ratio) between the two synthesis systems. Using ZSM-5—which has a broader crystallization window—as a model, we systematically investigated how M₂O/H₂O ratio influences the regulatory effect of “Na⁺/K⁺ ion switch”. As illustrated in Fig. S17, pure-phase ZSM-5 can be obtained with M₂O/H₂O = 0.0032–0.0112. At low alkalinity (M₂O/H₂O ≤ 0.0060), the K⁺ system produces smooth crystals (Fig. S17C–D), while the Na⁺ system yields nanocrystalline assemblies (Fig. S17G–H), consistent with what was revealed in our previous work.³⁸ It's noted that the size of nanocrystal units decreases with the increase of Na₂O/H₂O ratio from 0.0032 to 0.0060, which implies an enhanced order formation tendency in the precursor blocks. At a middle alkalinity (M₂O/H₂O = 0.0088), both K⁺ and Na⁺ systems yield transitional morphology (Fig. S17E and I), suggesting that the building-blocks in the two systems have become comparable. Conversely, at a high M₂O/H₂O ratio of 0.0112 (Fig. S17F and J), the Na⁺ system gives dense ZSM-5 crystals, while the K⁺ system yields ZSM-5 nanocrystalline assemblies—mirroring the behavior in the beta system reported here. This reversal highlights the system-specific response to alkali metal ions, which is critically modulated by the inorganic alkalinity.

Thereupon, the regulatory pathway of the “Na⁺/K⁺ ion switch” under different inorganic alkalinities is clarified (Scheme 2). (1) With increasing alkalinity and M⁺ concentration, the aggregation degree of the initial gel rises, and the precursors gradually transform from loose WLPs (Region I) to aggregated WLPs (Region II) and dense gels (Region III). (2) Only the moderately aggregated state favors the development of weak-crystalline nanoparticles that assemble via oriented attachment (Region II). Overly loose amorphous WLPs undergo rapid random attachment (Region I); while excessively dense gels, due to their difficulty in spontaneous adjustment, tend to follow the dissolution–recrystallization pathway instead (Region III). (3) Therefore, nanocrystalline assembly-type zeolite products arise only from moderately cross-linked gels (Region II), whereas overly loose or dense gels typically yield smooth and dense single crystals (or their derivatives), albeit through different building-blocks (Regions I & III). (4) The “Na⁺/K⁺ ion switch” could effectively regulate this process by tuning precursor charge and aggregation (Scheme 2B): under identical alkalinity, K⁺ leads to looser gels while Na⁺ favors denser ones, thus enabling direct switching of crystallization pathways and product mesostructure without altering other synthesis conditions.

Scheme 2 (A) Effects of inorganic base alkalinity on the gel state, building-blocks formation, and final zeolite morphology in the seed-induced synthesis system; (B) synergistic regulatory mechanism between the “Na⁺/K⁺ ion switch” strategy and the system's inorganic alkalinity.

Universality of the ion-switch strategy across zeolite frameworks

The elucidated mechanism implies that the “Na⁺/K⁺ ion switch” should enable fine-tuning of the mesostructure. To verify this, we employed mixed Na⁺–K⁺ feeds with varying molar ratios (K⁺ : Na⁺ = 1/3 to 3). Indeed, the morphology of the resulting beta zeolites transitioned continuously from relatively compact to clustered, loose architectures as the K⁺/Na⁺ ratio increased (Fig. 6A–C). This gradual transition in morphology and porosity underscores the continuous nature of the ion-switch effect. N₂-sorption analyses corroborate this trend, showing a systematic enhancement in textural properties (e.g., S_ext and V_meso) with higher K⁺ content (Fig. 6D and Table S1). These results demonstrate the capability of this strategy for customizing zeolite beta with tailored mesostructure and porosity for specific applications.

Fig. 6 (A)–(C) SEM (subscript 1) and TEM (subscript 2) images, (D) N₂-sorption isotherms and (inset) corresponding DFT pore size distribution, and (E) NH₃-TPD profiles for beta mesocrystals synthesized under different K⁺/Na⁺ molar ratios, where K : Na = (A) 1/3, (B) 1, and (C) 3, respectively.

To evaluate the universality of this strategy beyond beta and ZSM-5, we extended it to other seed-induced systems: ZSM-11 (MEL), zeolite L (LTL), and mordenite (MOR). For ZSM-11 (M₂O/H₂O = 0.0080), the K⁺ product shows a smoother surface than the Na⁺ product (Fig. S18), matching the regulatory trend at low alkalinity. For zeolite L (M₂O/H₂O = 0.01625), partial replacement of K⁺ with Na⁺ enlarged the nanodisc units and made the whole particle smoother and denser (Fig. S19), consistent with the effect of the “Na⁺/K⁺ ion switch” under high alkalinity. Additionally, in the synthesis system of mordenite via heterogeneous induction using calcined beta zeolite as seeds (M₂O/H₂O = 0.0150), partially replacing Na⁺ by K⁺ produced smaller, rougher crystals (Fig. S20), again in line with the high-alkalinity rule. Considering that many zeolites have specific requirements for the alkalinity range in their synthesis, the resulting products under traditional conditions tend to have a limited or monotonous morphology—either clustered or dense exclusively. In contrast, the “Na⁺/K⁺ ion switch” serves as a versatile mesostructure-engineering strategy, enabling controlled transitions between loose and dense states without altering other synthesis conditions, thereby broadening adaptability across diverse applications.

Hierarchical porosity-enhanced catalytic performance in macromolecular conversion.

A comprehensive acidity evaluation was conducted on H⁺-type beta-K/Na mesocrystals and their mixed-feed samples. NH₃-TPD and pyridine-IR spectra (Fig. 2D–E and 6E) revealed that all samples possess comparable total acid amounts, strength distributions (1429–1468 µmol g⁻¹ total acid sites and 499–544 µmol g⁻¹ strong acid sites, Table S2), and Brønsted/Lewis acid ratio (B/L = 1.74–1.82, Table S4),^48,49 consistent with their identical framework stoichiometry. This confirms that the “Na⁺/K⁺ ion switch” alters the mesostructure without affecting the framework composition or intrinsic acidity. However, the acid site accessibility, probed via non-aqueous potentiometric titration using a bulky molecule (2,2,6,6-tetramethylpiperidine of 8.91 Å, larger than the size of the *BEA zeolite micropore of 7.1 Å),^50,51 differed dramatically. The H⁺-type beta-K sample exhibited 102 µmol g⁻¹ of accessible acid sites, nearly 2.4 times that of the denser beta-Na sample (Fig. 2F and Table S4), which is a direct consequence of its abundant mesopores and larger external surface area.

The catalytic implication of this enhanced accessibility was demonstrated in the cracking of 1,3,5-triisopropylbenzene (TIPB, kinetic diameter 0.94 nm).⁵² While both catalysts achieved >99% conversion at 550 °C, a stark contrast emerged at lower temperatures (Fig. S21). At 250 °C, the conversion on beta-Na plummeted to ca. 25%, whereas beta-K maintained a remarkable conversion exceeding 90%. The result unequivocally demonstrates that the hollow, mesopore-rich architecture of beta-K provides superior accessibility to strong acid sites for bulky molecules, effectively mitigating intracrystalline diffusion limitations, thereby delivering superior performance in C–C bond cleavage reactions of bulky substrates.

We further assessed the catalysts in the cracking of low-density polyethylene (LDPE), a macromolecular model for plastic waste upcycling (Fig. 7).⁵³ In the isothermal cracking at 300 °C, beta-K achieved 90% conversion in merely 24 min, outperforming beta-Na (203 min) by nearly an order of magnitude (Fig. 7A). The turnover frequency (TOF) at 20% conversion for beta-K (0.524 min⁻¹) was more than three times higher than that of beta-Na (0.168 min⁻¹), while the mixed-feed samples fell in between, displaying a positive correlation with external surface area (Fig. 7C, left). Under the temperature-programmed conditions, beta-K reduced the characteristic cracking temperature (T₅₀) by ca. 70 °C compared to beta-Na (Fig. 7B).

Fig. 7 (A) The pyrolysis mass loss curves of LDPE at 300 °C. (B) The temperature-programmed (10 °C min⁻¹) weight loss curve of LDPE with different beta mesocrystals as catalysts. (C) TOF values of each catalyst at 20% LDPE conversion (left) and activation energy of LDPE cracking on different catalysts (right). (D) The product distribution measured via a flash cracking experiment at T₉₈.

Kinetic analysis with multiple heating-rate experiments (Fig. S22) provided deeper insight into the mechanism.⁵⁴ The apparent activation energy (E_a) for LDPE cracking over beta-K was initially about 10 kJ mol⁻¹ lower than that over beta-Na. More notably, as the reaction progressed, the E_a for C–C bond cleavage further decreased for beta-K, while it increased for beta-Na (Fig. 7C, right). This contrasting trend indicates a transition from diffusion-limited to kinetically controlled regimes within the hierarchical pores of beta-K, facilitated by its shorter intracrystalline diffusion paths. The open structure promotes product desorption and allows progressively cracked polymer fragments to access the internal strong acid sites more efficiently.^55–57 In contrast, the dense structure of beta-Na imposes persistent diffusion constraints, leading to catalyst deactivation and rising E_a.

Catalytic cycling tests further demonstrate the superior stability of H⁺ beta-K: after five consecutive LDPE cracking–regeneration cycles, H⁺ beta-K maintained lower T₅₀ shifts and higher activity than H⁺ beta-Na (Fig. S23). Post-reaction characterization studies revealed that the framework crystallinity, micropore volume, and acid site density of beta-K were largely preserved, whereas beta-Na showed noticeable loss of microporosity and acidity (Fig. S24 and Tables S5–S6). Thermogravimetric (TG) analysis under air confirmed that coke deposited on beta-K was more easily removed than on beta-Na (Fig. S25), consistent with the smaller activity drop observed during regeneration. These results indicate that beta-K not only provides higher intrinsic activity but also maintains structural integrity and regenerability, supporting its practical applicability for repeated LDPE cracking.

Finally, an in situ transient pyrolysis cell coupled with gas chromatography/mass spectrometry (PY-GC/MS) was used to evaluate product distribution (Fig. 7D). Without adding catalysts, the cracking products of LDPE were dominated by C₁₉₊ heavy aliphatic hydrocarbons (62.0%). By comparison, beta-K favored the best selectivity toward light hydrocarbons (C₁₈ and below). Specifically, the selectivity of C_2–4 liquefied petroleum gas (LPG) components reached 26.8% (vs. 16.2% for beta-Na), C_5–11 gasoline components accounted for 64.1%, and relatively more ArC_6–12 light aromatics (7.3%) were produced. These light aromatics can serve as anti-knock additives to improve the octane number of the resulting gasoline products. Overall, the beta-K mesocrystals—with hollow mulberry-like morphology and abundant accessible strong acid sites—enabled LDPE cracking and recycling at lower temperatures, giving mainly LPG and high-octane gasoline that can be directly used as chemical feedstocks and clean fuels, highlighting significant application potential.

Conclusions

In summary, we have demonstrated that the mesostructure of zeolite beta can be precisely engineered through a simple “Na⁺/K⁺ ion switch” strategy, which effectively steers the crystallization pathway by regulating precursor aggregation. Systematic investigation reveals that Na⁺ promotes dense gel formation and classical dissolution–recrystallization, yielding dense single crystals. In contrast, K⁺ maintains a moderate aggregation state, enabling the evolution of semi-crystalline nanoparticles and their subsequent oriented attachment into hierarchical, hollow nanocrystal assemblies. The catalytic superiority of the resulting beta-K mesocrystals in polyolefin cracking underscores the critical role of tailored mesoporosity in overcoming mass-transfer limitations for bulky molecules. Beyond zeolite beta, the successful extension of this strategy to multiple zeolite frameworks (ZSM-5, ZSM-11, zeolite L and mordenite) confirms its broad universality. The synergistic role of inorganic alkalinity in directing the ion-specific switching behavior has been clearly deciphered, establishing a general model for predictable mesostructure control. Our findings not only provide deep mechanistic insights into non-classical crystallization pathways in zeolite synthesis but also establish a green, scalable, and economically viable route for the target-oriented design of advanced catalysts. This work opens new avenues for applications in sustainable chemical processes, particularly in the catalytic upcycling of plastic wastes.

Experimental

Materials

Chemicals used for beta mesocrystal synthesis via seed-induced route: silica sol (LUDOX HS-40, 40 wt% SiO₂ in H₂O), aluminum sulfate (Al₂(SO₄)₃·18H₂O, AR, Shanghai Chemical Co.), tetraethylammonium hydroxide (TEAOH, 25 wt% in H₂O, Aladdin), potassium hydroxide (KOH, semiconductor grade, Aladdin), sodium hydroxide (NaOH, AR, Shanghai Chemical Co.), and deionized water. For beta seed synthesis: fumed silica (Aerosil400, Shanghai Chlorine Alkali Industry), aluminum foil (SP, Aladdin), TEAOH (25 wt% in H₂O, Aladdin), and deionized water. For etching experiment: NaOH (AR, Shanghai Chemical Co.) and deionized water. For catalysis: low-density polyethylene (LDPE, Alfa-Aesar) and 1,3,5-triisopropylbenzene (TIPB, ≥ 95%, Aladdin).

Synthesis of samples

The beta seed (ca. 220 nm) was synthesized following a previous report,⁵⁹ with suspension centrifuged and redispersed in deionized water (15 wt%). The PXRD and electron microscopy images of the as-synthesized beta seed are shown in Fig. S1. The seed should be used without drying or calcination to preserve its hydrophilicity. The beta-K and beta-Na samples were prepared via a seed-induced route: a mixture with the composition n_SiO2/n_Al2O3/n_TEA2O/n_M2O/n_H2O = 100/3.333/5/20/2000 (M = K or Na) was prepared, followed by the addition of pre-synthesized beta seed dispersion. Specifically, after mixing 20 wt% NaOH or KOH solution and 25 wt% TEAOH solution, 40 wt% silica sol was slowly added under stirring, followed by dropwise addition of 10 wt% Al₂(SO₄)₃·18H₂O solution and the remaining water. After 30 min aging, the seed dispersion was added dropwise (7 wt% of the total SiO₂ weight), and the gel was stirred at room temperature for another 3.5 h before transferring into a Teflon-lined autoclave and heated at 140 °C under static conditions for 60 h. After hydrothermal treatment, solid products were separated by centrifugation (12 000 rpm for 10 min; supernatant should also be retained for further analysis) and washed several times with deionized water. To investigate the crystallization mechanism, intermediates were extracted at the pre-determined time and denoted as beta-K/Na-t, where t (h) refers to the heating time. In order to preserve the original morphology, especially for the intermediates, all washed samples were freeze-dried under vacuum at −50 °C to minimize structure change. For calcination, dried products were heated at 550 °C for 6 h to remove templates before conducting Raman, N₂-sorption analysis, alkali treatment and NH₄⁺ ion-exchange. The typical samples for acidity evaluation were ion-exchanged three times with NH₄Cl solution (5 wt%) at 90 °C for 3 h, followed by calcination at 550 °C for 6 h.

Verification and expansion experiments: synthesis without seeds was performed using the same procedure, except that no beta seed dispersion was added, and the heating time at 140 °C was extended to 7 days. Synthesis without TEAOH was conducted using the same seed-induced procedure but without adding TEAOH solution (composition adjusted to n_SiO2/n_Al2O3/n_M2O/n_H2O = 100/3.333/25/2000 to maintain the total alkalinity). And transition samples of beta-K and beta-Na were synthesized under the same total basicity, water content and Si/Al ratio at 140 °C for 60 h, with the molar ratios of K⁺/Na⁺ set to 1/3, 1, and 3, respectively. Furthermore, an alkali etching experiment⁵⁸ was performed by dispersing 0.5 g of calcined product in 20 g of 0.2 mol L⁻¹ NaOH solution at 65 °C for 1 h under stirring. Additionally, the synthesis details of ZSM-5 (MFI), ZSM-11 (MEL), mordenite (MOR) and zeolite L (LTL) are listed in the SI.

Characterization

The powder X-ray diffraction (PXRD) experiments were conducted on a Bruker D2 diffractometer (Cu K_α, 30 kV, 10 mA). The sample morphologies were observed by field-emission scanning electron microscopy (FE-SEM, Hitachi S-4800). Further structural analysis was carried out using field-emission transmission electron microscopy (FE-TEM), selected area electron diffraction (SAED), EDS mapping, and high-angle annular dark-field scanning TEM (HAADF-STEM) on a Tecnai G2 F20 S-Twin instrument. Physical parameters of samples were measured by N₂-sorption at 77 K and Ar-sorption at 87 K (Quantachrome iQ-2) after outgassing at 573 K for 8 h. Fourier-transform infrared spectra (FT-IR) were collected using a Bruker Invenio S spectrometer after outgassing under vacuum at 450 °C for 2 h. ²⁷Al, ²⁹Si MAS NMR and ¹H–²⁹Si cross-polarization (CP) NMR were determined on a Bruker 400WB AVANCE III spectrometer. Inductively coupled plasma-atomic emission spectrometry (ICP-AES, iCAP 7400) was used to analyze the concentration changes in the supernatant, and elemental composition of solid samples was measured after hydrofluoric acid dissolution. UV-Raman spectra were obtained from a Horiba LabRAM HR Evolution confocal microscope with a 325 nm laser. Small-angle X-ray scattering (SAXS) measurements for the dried samples were performed on a Xeuss 2.0 SAXS system (Xenocs, France) using Metaljet Ga K_α radiation. The acidic properties were measured via NH₃ temperature-programmed desorption (NH₃-TPD) on a Micromeritics AutoChem II 2920. Moreover, pyridine adsorption infrared spectroscopy (pyridine-IR) was used to investigate the acid site properties at 300 °C^48,49 on a Bruker Invenio S instrument, and the number of accessible acid sites was determined by nonaqueous titration using 2,2,6,6-tetramethylpiperidine (0.01 mol L⁻¹ in acetonitrile) on a potentiometric titration meter (ZDJ-5, Shanghai Leici).^50,51 For dynamic light scattering (DLS) and zeta potential measurements (both on Nano-ZS90 zetasizer), the aged gel mixtures were dispersed (ca. 0.2 wt%) in their corresponding centrifuged supernatants prior to analysis.

Catalytic tests

LDPE cracking reaction was carried out in a thermogravimetric reactor (SDT650). Typically, 16 mg of LDPE was mixed in an alumina crucible with H⁺-type zeolite catalyst at a mass ratio of 8 : 1. N₂ was used as the carrier gas (100 mL min⁻¹) to purge air in advance. The reaction was performed under different temperature-rising conditions, such as heating from room temperature to 600 °C at a rate of 5/10/20 °C min⁻¹, or heating to 300 °C at 20 °C min⁻¹ followed by isothermal reaction at 300 °C for several hours. Quantitative evaluation of catalytic cracking activity, together with catalytic cycling tests and post-reaction physicochemical characterization of the catalysts, is provided in the SI. An in situ transient micro-pyrolysis flash cracking device was used to analyze the product distribution. The temperature at which 98% weight loss of LDPE occurred (T₉₈), as determined in the aforementioned 10 °C min⁻¹ temperature-programmed experiments, was selected to perform transient thermal flash cracking. The product distribution was analyzed using a GC/MS system (PY-GC/MS, Agilent 7890A-5975C, HP-5MS capillary column).

TIPB cracking was conducted in a tandem μ-reactor (Frontier Lab, Rx-3050TR) to evaluate the catalytic performance of the beta zeolite. Typically, 30 mg of H⁺-type zeolite was placed in a quartz tube μ-reactor and activated under N₂ flow (45 mL min⁻¹) at 550 °C for 2 h. TIPB was injected into the reaction system via pulse injection (2.0 µL). The products were analyzed online using a Shimadzu GC-2030 chromatograph equipped with a hydrogen ion flame detector (FID) and HP-5 capillary column.

Author contributions

Zhaoqi Ye: conceptualization, methodology, investigation, data curation, formal analysis, writing – original draft, visualization, funding acquisition. Yifan Zhang: investigation, formal analysis, validation. Kexin Yan: methodology, investigation, formal analysis. Hongbin Zhang: funding acquisition, conceptualization, formal analysis, visualization, writing – review & editing. Zhengmin Yu: investigation, data curation. Zhizheng Sheng: investigation, validation. Ke Du: formal analysis, data curation. Yahong Zhang: funding acquisition, formal analysis, writing – review & editing. Yi Tang: funding acquisition, resources, formal analysis, writing – review & editing, supervision, project administration.

Conflicts of interest

The authors declare no conflicts of interest.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information: details of synthesis, characterization and catalytic tests; PXRD, SEM, TEM, Ar- and N₂-sorption, ²⁷Al, ²⁹Si and ¹H–²⁹Si NMR, UV-Raman, DLS, zeta potential, SAXS, optical photograph, elemental analysis and NH₃-TPD of samples; catalytic performance for zeolite beta mesocrystals. See DOI: https://doi.org/10.1039/d5qm00905g.

Acknowledgements

This work was supported by the National Key R&D Program of China (No. 2023YFA1507602), the NSFC (No. 22402036, 22088101, and 22175040) and the Science and Technology Commission of Shanghai Municipality (No. 2024ZDSYS02).

References

1. H. Xu and P. Wu, New progress in zeolite synthesis and catalysis, Natl. Sci. Rev., 2022, 9, nwac045.
2. Z. X. Qin, S. Zeng, G. Melinte, T. Bucko, M. Badawi, Y. F. Shen, J. P. Gilson, O. Ersen, Y. X. Wei, Z. M. Liu, X. M. Liu, Z. F. Yan, S. T. Xu, V. Valtchev and S. Mintova, Understanding the fundamentals of microporosity upgrading in zeolites: increasing diffusion and catalytic performances, Adv. Sci., 2021, 8, e2100001.
3. M. Shamzhy, M. Opanasenko, P. Concepcion and A. Martinez, New trends in tailoring active sites in zeolite-based catalysts, Chem. Soc. Rev., 2019, 48, 1095–1149.
4. B. J. Zeng, S. M. Wu, M. Antonietti, C. Janiak, O. Terasaki and X. Y. Yang, Hierarchy design of zeolite mesocrystals, Chem. Eng. J., 2025, 505, 159282.
5. A. J. Mallette, S. Seo and J. D. Rimer, Synthesis strategies and design principles for nanosized and hierarchical zeolites, Nat. Synth, 2022, 1, 521–534.
6. W. Schwieger, A. G. Machoke, T. Weissenberger, A. Inayat, T. Selvam, M. Klumpp and A. Inayat, Hierarchy concepts: classification and preparation strategies for zeolite containing materials with hierarchical porosity, Chem. Soc. Rev., 2016, 45, 3353–3376.
7. J. Jeong, K. Cho, S. H. Kim, D. Lee, J. I. Park, D. Bae, K. Cho and J. C. Kim, Zeolite beta nanosponge with high acidity on mesopores for efficient tributyrin transesterification, Fuel, 2025, 402, 135965.
8. S. Fernandez, M. L. Ostraat and K. Zhang, Toward rational design of hierarchical beta zeolites: an overview and beyond, AIChE J., 2020, 66, e16943.
9. B. Xie, J. W. Song, L. M. Ren, Y. Y. Ji, J. X. Li and F. S. Xiao, Organotemplate-free and fast route for synthesizing beta zeolite, Chem. Mater., 2008, 20, 4533–4535.
10. Y. Ma, X. M. Tang, J. Y. Hu, Y. H. Ma, W. Chen, Z. Q. Liu, S. C. Han, C. Xu, Q. M. Wu, A. M. Zheng, L. F. Zhu, X. J. Meng and F. S. Xiao, Design of a small organic template for the synthesis of self-pillared pentasil zeolite nanosheets, J. Am. Chem. Soc., 2022, 144, 6270–6277.
11. A. Korde, B. Min, E. Kapaca, O. Knio, I. Nezam, Z. Y. Wang, J. Leisen, X. Y. Yin, X. Y. Zhang, D. S. Sholl, X. D. Zou, T. Willhammar, C. W. Jones and S. Nair, Single-walled zeolitic nanotubes, Science, 2022, 375, 62–66.
12. X. L. Jia, J. G. Jiang, S. H. Zou, L. Han, H. Y. Zhu, Q. Zhang, Y. Ma, P. Luo, P. Wu, A. Mayoral, X. B. Han, J. Cheng and S. A. Che, Library creation of ultrasmall multi-metallic nanoparticles confined in mesoporous MFI zeolites, Angew. Chem., Int. Ed., 2021, 60, 14571–14577.
13. Y. J. Zhang and S. A. Che, pi–pi Interactions between aromatic groups in amphiphilic molecules: directing hierarchical growth of porous zeolites, Angew. Chem., Int. Ed., 2020, 59, 50–60.
14. K. Yan, Y. Zhao, C. Zhao, H. Li, Z. Ye, X. Yang, Y. Zhang, H. Zhang and Y. Tang, K⁺ promoted fabrication of nanoneedle low-silicon ZSM-48 mesocrystal, Chem. Synth., 2024, 4, 38.
15. M. H. Sun, L. H. Chen, S. Yu, Y. Li, X. G. Zhou, Z. Y. Hu, Y. H. Sun, Y. Xu and B. L. Su, Micron-sized zeolite beta single crystals featuring intracrystal interconnected ordered macro-meso-microporosity displaying superior catalytic performance, Angew. Chem., Int. Ed., 2020, 59, 19582–19591.
16. A. Azhati, S. H. Xie, W. W. Wang, A. A. Elzatahry, Y. E. Yan, J. Zhou, D. Al-Dhayan, Y. H. Zhang, Y. Tang and D. Y. Zhao, Ordered, highly zeolitized mesoporous aluminosilicates produced by a gradient acidic assembly growth strategy in a mixed template system, Chem. Mater., 2016, 28, 4859–4866.
17. M. H. Sun, J. Zhou, Z. Y. Hu, L. H. Chen, L. Y. Li, Y. D. Wang, Z. K. Xie, S. Turner, G. Van Tendeloo, T. Hasan and B. L. Su, Hierarchical zeolite single-crystal reactor for excellent catalytic efficiency, Matter, 2020, 3, 1226–1245.
18. X. C. Zhu, Y. Gao, H. R. Chen, M. W. Jiang, X. Y. Wang, C. X. Miao, Y. F. Shen, Y. G. Ji, Z. X. Qin, Z. J. Wu, W. Y. Song, C. M. Xu and B. J. Shen, Constructing hierarchical zeolites with highly complete framework via controlled desilication, Angew. Chem., Int. Ed., 2024, 63, e202411446.
19. Z. X. Qin, Z. C. You, K. N. Bozhilov, S. K. Kolev, W. Yang, Y. F. Shen, X. Jin, J. P. Gilson, S. Mintova, G. N. Vayssilov and V. Valtchev, Dissolution behavior and varied mesoporosity of zeolites by NH₄F etching, Chem. - Eur. J., 2022, 28, e202104339.
20. J. J. De Yoreo, P. U. P. A. Gilbert, N. A. J. M. Sommerdijk, R. L. Penn, S. Whitelam, D. Joester, H. Z. Zhang, J. D. Rimer, A. Navrotsky, J. F. Banfield, A. F. Wallace, F. M. Michel, F. C. Meldrum, H. Colfen and P. M. Dove, Crystallization by particle attachment in synthetic, biogenic, and geologic environments, Science, 2015, 349, aaa6760.
21. A. J. Mallette, K. Shilpa and J. D. Rimer, The current understanding of mechanistic pathways in zeolite crystallization, Chem. Rev., 2024, 124, 3416–3493.
22. Z. Z. Sheng, H. Li, K. Du, L. Gao, J. Ju, Y. H. Zhang and Y. Tang, Observing a zeolite nucleus (subcrystal) with a uniform framework structure and its oriented attachment without single-molecule addition, Angew. Chem., Int. Ed., 2021, 60, 13444–13451.
23. M. Kumar, R. Li and J. D. Rimer, Assembly and evolution of amorphous precursors in zeolite L crystallization, Chem. Mater., 2016, 28, 1714–1727.
24. K. Du, X. Zhang, T. Y. He, W. Shen, H. L. Xu, Y. Tang, X. M. Cao, Z. Huang and Y. H. Zhang, Dynamic evolution of structural ordering and aluminum redistribution during ZSM-5 zeolite crystallization, Angew. Chem., Int. Ed., 2025, 64, e202507223.
25. Q. Zhang, J. Y. Li, X. X. Wang, G. Y. He, L. Li, J. Xu, D. H. Mei, O. Terasaki and J. H. Yu, Silanol-engineered nonclassical growth of zeolite nanosheets from oriented attachment of amorphous protozeolite nanoparticles, J. Am. Chem. Soc., 2023, 145, 21231–21241.
26. K. N. Olafson, R. Li, B. G. Alamani and J. D. Rimer, Engineering crystal modifiers: bridging classical and nonclassical crystallization, Chem. Mater., 2016, 28, 8453–8465.
27. C. S. Cundy and P. A. Cox, The hydrothermal synthesis of zeolites: precursors, intermediates and reaction mechanism, Microporous Mesoporous Mater., 2005, 82, 1–78.
28. D. Verboekend, G. Vile and J. Perez-Ramirez, Synthesis strategies in the search for hierarchical zeolites, Chem. Soc. Rev., 2013, 42, 4004–4035.
29. H. Dai, Y. F. Shen, T. M. Yang, C. S. Lee, D. L. Fu, A. Agarwal, T. T. Le, M. Tsapatsis, J. C. Palmer, B. M. Weckhuysen, P. J. Dauenhauer, X. D. Zou and J. D. Rimer, Finned zeolite catalysts, Nat. Mater., 2020, 19, 1074–1080.
30. M. K. Choudhary, R. Jain and J. D. Rimer, In situ imaging of two-dimensional surface growth reveals the prevalence and role of defects in zeolite crystallization, Proc. Natl. Acad. Sci. U. S. A., 2020, 117, 28632–28639.
31. M. Kumar, M. K. Choudhary and J. D. Rimer, Transient modes of zeolite surface growth from 3D gel-like islands to 2D single layers, Nat. Commun., 2018, 9, 2129.
32. H. B. Zhang, H. X. Zhang, Y. Zhao, Z. P. Shi, Y. H. Zhang and Y. Tang, Seeding bundlelike MFI zeolite mesocrystals: a dynamic, nonclassical crystallization via epitaxially anisotropic growth, Chem. Mater., 2017, 29, 9247–9255.
33. Z. Q. Ye, Y. Zhao, H. B. Zhang, Y. H. Zhang and Y. Tang, Co-hydrolysis and seed-induced synthesis of basic mesoporous ZSM-5 zeolites with enhanced catalytic performance, Chem. - Eur. J., 2020, 26, 6147–6157.
34. Y. L. Liu, Z. Q. Han, M. Yi, Z. Y. Zhang, Z. F. You, X. L. Liu, Q. Zhao and B. Y. Li, Ultrafast sonochemical synthesis of imine-linked porous organic cages with high surface area for gas adsorption, Mater. Horiz., 2025, 12, 8147–8152.
35. J. Chen, W. M. Hua, Y. Xiao, Q. S. Huo, K. K. Zhu and X. G. Zhou, Tailoring the structure of hierarchically porous zeolite beta through modified orientated attachment growth in a dry gel system, Chem. - Eur. J., 2014, 20, 14744–14755.
36. K. Iyoki, K. Itabashi and T. Okubo, Seed-assisted, one-pot synthesis of hollow zeolite beta without using organic structure-directing agents, Chem. - Asian J., 2013, 8, 1419–1427.
37. Z. Ma, Q. Zhang, L. Li, M. Y. Chen, J. Y. Li and J. H. Yu, Steam-assisted crystallization of highly dispersed nanosized hierarchical zeolites from solid raw materials and their catalytic performance in lactide production, Chem. Sci., 2022, 13, 8052–8059.
38. Z. Q. Ye, Y. Zhao, H. B. Zhang, Z. P. Shi, H. Li, X. Yang, L. Wang, L. T. Kong, C. N. Zhang, Z. Z. Sheng, Y. H. Zhang and Y. Tang, Mesocrystal morphology regulation by “alkali metals ion switch”: re-examining zeolite nonclassical crystallization in seed-induced process, J. Colloid Interface Sci., 2022, 608, 1366–1376.
39. F. Lin, Z. Q. Ye, L. T. Kong, P. Liu, Y. H. Zhang, H. B. Zhang and Y. Tang, Facile morphology and porosity regulation of zeolite ZSM-5 mesocrystals with synergistically enhanced catalytic activity and shape selectivity, Nanomaterials, 2022, 12, 1601.
40. L. Wang, N. N. Yan, X. N. Liu, X. B. Zhao, M. K. Shen, L. F. Liu, P. Tian, P. Guo and Z. M. Liu, Unraveling the twin and tunability of the crystal domain sizes in the medium-pore zeolite ZSM-57 by electron crystallography, Chem. - Eur. J., 2019, 25, 1029–1036.
41. Z. Ye, L. Kong, Y. Zhao, C. Zhang, X. Yang, K. Yan, Y. Zhang, H. Zhang and Y. Tang, Alkalinity-controlled zeolite nucleation and growth: ultrafast synthesis of total-morphology zeolite L mesocrystals and adsorption evaluation, Chem. Synth., 2022, 2, 20.
42. M. Q. Tong, D. L. Zhang, L. K. Zhu, J. Xu, F. Deng, R. R. Xu and W. F. Yan, An elaborate structure investigation of the chiral polymorph A-enriched zeolite beta, CrystEngComm, 2016, 18, 1782–1789.
43. M. Thommes, K. Kaneko, A. V. Neimark, J. P. Olivier, F. Rodriguez-Reinoso, J. Rouquerol and K. S. W. Sing, Physisorption of gases, with special reference to the evaluation of surface area and pore size distribution (IUPAC Technical Report), Pure Appl. Chem., 2015, 87, 1051–1069.
44. E. Dib, E. B. Clatworthy, H. Cruchade, I. C. Medeiros-Costa, N. Nesterenko, J. P. Gilson and S. Mintova, Exploration, explanation and exploitation of hydroxyls in zeolites, Natl. Sci. Rev., 2022, 9, nwac081.
45. G. Harvey and L. S. D. Glasser, Structure and properties of aluminosilicate solutions and gels, ACS Symp. Ser., 1989, 398, 49–65.
46. Z. X. Qin, J. P. Gilson and V. Valtchev, Mesoporous zeolites by fluoride etching, Curr. Opin. Chem. Eng., 2015, 8, 1–6.
47. S. Abello, A. Bonilla and J. Perez-Ramirez, Mesoporous ZSM-5 zeolite catalysts prepared by desilication with organic hydroxides and comparison with NaOH leaching, Appl. Catal., A, 2009, 364, 191–198.
48. M. Guisnet, P. Ayrault and J. Datka, Acid properties of dealuminated mordenites studied by IR spectroscopy. 2. Concentration, acid strength and heterogeneity of OH groups, Pol. J. Chem., 1997, 71, 1455–1461.
49. G. Li, L. Gao, Z. Z. Sheng, Y. L. Zhan, C. Y. Zhang, J. Ju, Y. H. Zhang and Y. Tang, A Zr-Al-Beta zeolite with open Zr(IV) sites: an efficient bifunctional Lewis-Brønsted acid catalyst for a cascade reaction, Catal. Sci. Technol., 2019, 9, 4055–4065.
50. X. Yang, B. X. Gao, W. Y. Li, K. G. Hou, K. X. Yan, T. L. Yan, S. N. Wang, Y. H. Zhang, L. Wang and Y. Tang, Mechanism investigation of acid-tailored USY zeolites via ion exchange in polypropylene cracking, Chem. Catal., 2024, 4, 100947.
51. H. B. Zhang, Y. C. Ma, K. S. Song, Y. H. Zhang and Y. Tang, Nano-crystallite oriented self-assembled ZSM-5 zeolite and its LDPE cracking properties: Effects of accessibility and strength of acid sites, J. Catal., 2013, 302, 115–125.
52. W. J. Dai, L. N. Zhang, R. Z. Liu, G. J. Wu, N. J. Guan and L. D. Li, Plate-Like ZSM-5 Zeolites as Robust Catalysts for the Cracking of Hydrocarbons, ACS Appl. Mater. Interfaces, 2022, 14, 11415–11424.
53. M. Y. Chu, Y. Liu, X. X. Lou, Q. Zhang and J. X. Chen, Rational Design of Chemical Catalysis for Plastic Recycling, ACS Catal., 2022, 12, 4659–4679.
54. A. O. S. Silva, M. J. B. Souza, A. M. G. Pedrosa, A. C. F. Coriolano, V. J. Fernandes and A. S. Araujo, Development of HZSM-12 zeolite for catalytic degradation of high-density polyethylene, Microporous Mesoporous Mater., 2017, 244, 1–6.
55. J. D. Duan, W. Chen, C. T. Wang, L. Wang, Z. Q. Liu, X. F. Yi, W. Fang, H. Wang, H. Wei, S. D. Xu, Y. W. Yang, Q. W. Yang, Z. B. Bao, Z. G. Zhang, Q. L. Ren, H. Zhou, X. D. Qin, A. M. Zheng and F. S. Xiao, Coking-resistant polyethylene upcycling modulated by zeolite micropore diffusion, J. Am. Chem. Soc., 2022, 144, 14269–14277.
56. K. Du, Z. Z. Sheng, H. Li, L. Gao, T. L. Yan, J. Y. Yu, D. Pan, Z. C. Tan, S. H. Xie, J. Ju, Y. H. Zhang and Y. Tang, Ultrasmall zeolite subcrystal catalyst devoid of intracrystalline diffusion limitation for bulk molecule conversion, ACS Appl. Nano Mater., 2023, 7, 1146–1155.
57. B. Reiprich, K. A. Tarach, K. Pyra, G. Grzybek and K. Góra-Marek, High-silica layer-like zeolites Y from seeding-free synthesis and their catalytic performance in low-density polyethylene cracking, ACS Appl. Mater. Interfaces, 2022, 14, 6667–6679.
58. M. A. Camblor, A. Corma, A. Mifsud, J. Perez-Pariente and S. Valencia, Synthesis of nanocrystalline zeolite beta in the absence of alkali metal cations, Stud. Surf. Sci. Catal., 1997, 105, 341–348.
59. J. C. Groen, S. Abelló, L. A. Villaescusa and J. Pérez-Ramírez, Mesoporous beta zeolite obtained by desilication, Microporous Mesoporous Mater., 2008, 114, 93–102.

---